This commit was manufactured by cvs2svn to create tag
[linux-2.6.git] / arch / x86_64 / kernel / ptrace.c
index fb883d3..98f9c2e 100644 (file)
@@ -17,8 +17,6 @@
 #include <linux/user.h>
 #include <linux/security.h>
 #include <linux/audit.h>
-#include <linux/seccomp.h>
-#include <linux/signal.h>
 
 #include <asm/uaccess.h>
 #include <asm/pgtable.h>
@@ -64,12 +62,6 @@ static inline unsigned long get_stack_long(struct task_struct *task, int offset)
        return (*((unsigned long *)stack));
 }
 
-static inline struct pt_regs *get_child_regs(struct task_struct *task)
-{
-       struct pt_regs *regs = (void *)task->thread.rsp0;
-       return regs - 1;
-}
-
 /*
  * this routine will put a word on the processes privileged stack. 
  * the offset is how far from the base addr as stored in the TSS.  
@@ -87,130 +79,6 @@ static inline long put_stack_long(struct task_struct *task, int offset,
        return 0;
 }
 
-#define LDT_SEGMENT 4
-
-unsigned long convert_rip_to_linear(struct task_struct *child, struct pt_regs *regs)
-{
-       unsigned long addr, seg;
-
-       addr = regs->rip;
-       seg = regs->cs & 0xffff;
-
-       /*
-        * We'll assume that the code segments in the GDT
-        * are all zero-based. That is largely true: the
-        * TLS segments are used for data, and the PNPBIOS
-        * and APM bios ones we just ignore here.
-        */
-       if (seg & LDT_SEGMENT) {
-               u32 *desc;
-               unsigned long base;
-
-               down(&child->mm->context.sem);
-               desc = child->mm->context.ldt + (seg & ~7);
-               base = (desc[0] >> 16) | ((desc[1] & 0xff) << 16) | (desc[1] & 0xff000000);
-
-               /* 16-bit code segment? */
-               if (!((desc[1] >> 22) & 1))
-                       addr &= 0xffff;
-               addr += base;
-               up(&child->mm->context.sem);
-       }
-       return addr;
-}
-
-static int is_at_popf(struct task_struct *child, struct pt_regs *regs)
-{
-       int i, copied;
-       unsigned char opcode[16];
-       unsigned long addr = convert_rip_to_linear(child, regs);
-
-       copied = access_process_vm(child, addr, opcode, sizeof(opcode), 0);
-       for (i = 0; i < copied; i++) {
-               switch (opcode[i]) {
-               /* popf */
-               case 0x9d:
-                       return 1;
-
-                       /* CHECKME: 64 65 */
-
-               /* opcode and address size prefixes */
-               case 0x66: case 0x67:
-                       continue;
-               /* irrelevant prefixes (segment overrides and repeats) */
-               case 0x26: case 0x2e:
-               case 0x36: case 0x3e:
-               case 0x64: case 0x65:
-               case 0xf0: case 0xf2: case 0xf3:
-                       continue;
-
-               /* REX prefixes */
-               case 0x40 ... 0x4f:
-                       continue;
-
-                       /* CHECKME: f0, f2, f3 */
-
-               /*
-                * pushf: NOTE! We should probably not let
-                * the user see the TF bit being set. But
-                * it's more pain than it's worth to avoid
-                * it, and a debugger could emulate this
-                * all in user space if it _really_ cares.
-                */
-               case 0x9c:
-               default:
-                       return 0;
-               }
-       }
-       return 0;
-}
-
-static void set_singlestep(struct task_struct *child)
-{
-       struct pt_regs *regs = get_child_regs(child);
-
-       /*
-        * Always set TIF_SINGLESTEP - this guarantees that
-        * we single-step system calls etc..  This will also
-        * cause us to set TF when returning to user mode.
-        */
-       set_tsk_thread_flag(child, TIF_SINGLESTEP);
-
-       /*
-        * If TF was already set, don't do anything else
-        */
-       if (regs->eflags & TRAP_FLAG)
-               return;
-
-       /* Set TF on the kernel stack.. */
-       regs->eflags |= TRAP_FLAG;
-
-       /*
-        * ..but if TF is changed by the instruction we will trace,
-        * don't mark it as being "us" that set it, so that we
-        * won't clear it by hand later.
-        *
-        * AK: this is not enough, LAHF and IRET can change TF in user space too.
-        */
-       if (is_at_popf(child, regs))
-               return;
-
-       child->ptrace |= PT_DTRACE;
-}
-
-static void clear_singlestep(struct task_struct *child)
-{
-       /* Always clear TIF_SINGLESTEP... */
-       clear_tsk_thread_flag(child, TIF_SINGLESTEP);
-
-       /* But touch TF only if it was set by us.. */
-       if (child->ptrace & PT_DTRACE) {
-               struct pt_regs *regs = get_child_regs(child);
-               regs->eflags &= ~TRAP_FLAG;
-               child->ptrace &= ~PT_DTRACE;
-       }
-}
-
 /*
  * Called by kernel/ptrace.c when detaching..
  *
@@ -218,7 +86,11 @@ static void clear_singlestep(struct task_struct *child)
  */
 void ptrace_disable(struct task_struct *child)
 { 
-       clear_singlestep(child);
+       long tmp;
+
+       clear_tsk_thread_flag(child, TIF_SINGLESTEP);
+       tmp = get_stack_long(child, EFL_OFFSET) & ~TRAP_FLAG;
+       put_stack_long(child, EFL_OFFSET, tmp);
 }
 
 static int putreg(struct task_struct *child,
@@ -257,13 +129,13 @@ static int putreg(struct task_struct *child,
                        value &= 0xffff;
                        return 0;
                case offsetof(struct user_regs_struct,fs_base):
-                       if (value >= TASK_SIZE)
-                               return -EIO;
+                       if (!((value >> 48) == 0 || (value >> 48) == 0xffff))
+                               return -EIO; 
                        child->thread.fs = value;
                        return 0;
                case offsetof(struct user_regs_struct,gs_base):
-                       if (value >= TASK_SIZE)
-                               return -EIO;
+                       if (!((value >> 48) == 0 || (value >> 48) == 0xffff))
+                               return -EIO; 
                        child->thread.gs = value;
                        return 0;
                case offsetof(struct user_regs_struct, eflags):
@@ -277,11 +149,6 @@ static int putreg(struct task_struct *child,
                                return -EIO;
                        value &= 0xffff;
                        break;
-               case offsetof(struct user_regs_struct, rip):
-                       /* Check if the new RIP address is canonical */
-                       if (value >= TASK_SIZE)
-                               return -EIO;
-                       break;
        }
        put_stack_long(child, regno - sizeof(struct pt_regs), value);
        return 0;
@@ -382,7 +249,7 @@ asmlinkage long sys_ptrace(long request, long pid, unsigned long addr, long data
                        break;
 
                switch (addr) { 
-               case 0 ... sizeof(struct user_regs_struct) - sizeof(long):
+               case 0 ... sizeof(struct user_regs_struct):
                        tmp = getreg(child, addr);
                        break;
                case offsetof(struct user, u_debugreg[0]):
@@ -427,7 +294,7 @@ asmlinkage long sys_ptrace(long request, long pid, unsigned long addr, long data
                        break;
 
                switch (addr) { 
-               case 0 ... sizeof(struct user_regs_struct) - sizeof(long):
+               case 0 ... sizeof(struct user_regs_struct)
                        ret = putreg(child, addr, data);
                        break;
                /* Disallows to set a breakpoint into the vsyscall */
@@ -472,10 +339,11 @@ asmlinkage long sys_ptrace(long request, long pid, unsigned long addr, long data
                }
                break;
        case PTRACE_SYSCALL: /* continue and stop at next (return from) syscall */
-       case PTRACE_CONT:    /* restart after signal. */
+       case PTRACE_CONT: { /* restart after signal. */
+               long tmp;
 
                ret = -EIO;
-               if (!valid_signal(data))
+               if ((unsigned long) data > _NSIG)
                        break;
                if (request == PTRACE_SYSCALL)
                        set_tsk_thread_flag(child,TIF_SYSCALL_TRACE);
@@ -483,11 +351,14 @@ asmlinkage long sys_ptrace(long request, long pid, unsigned long addr, long data
                        clear_tsk_thread_flag(child,TIF_SYSCALL_TRACE);
                clear_tsk_thread_flag(child, TIF_SINGLESTEP);
                child->exit_code = data;
-               /* make sure the single step bit is not set. */
-               clear_singlestep(child);
+       /* make sure the single step bit is not set. */
+               tmp = get_stack_long(child, EFL_OFFSET);
+               tmp &= ~TRAP_FLAG;
+               put_stack_long(child, EFL_OFFSET,tmp);
                wake_up_process(child);
                ret = 0;
                break;
+       }
 
 #ifdef CONFIG_IA32_EMULATION
                /* This makes only sense with 32bit programs. Allow a
@@ -524,28 +395,41 @@ asmlinkage long sys_ptrace(long request, long pid, unsigned long addr, long data
  * perhaps it should be put in the status that it wants to 
  * exit.
  */
-       case PTRACE_KILL:
+       case PTRACE_KILL: {
+               long tmp;
+
                ret = 0;
                if (child->exit_state == EXIT_ZOMBIE)   /* already dead */
                        break;
                clear_tsk_thread_flag(child, TIF_SINGLESTEP);
                child->exit_code = SIGKILL;
                /* make sure the single step bit is not set. */
-               clear_singlestep(child);
+               tmp = get_stack_long(child, EFL_OFFSET) & ~TRAP_FLAG;
+               put_stack_long(child, EFL_OFFSET, tmp);
                wake_up_process(child);
                break;
+       }
+
+       case PTRACE_SINGLESTEP: {  /* set the trap flag. */
+               long tmp;
 
-       case PTRACE_SINGLESTEP:    /* set the trap flag. */
                ret = -EIO;
-               if (!valid_signal(data))
+               if ((unsigned long) data > _NSIG)
                        break;
                clear_tsk_thread_flag(child,TIF_SYSCALL_TRACE);
-               set_singlestep(child);
+               if ((child->ptrace & PT_DTRACE) == 0) {
+                       /* Spurious delayed TF traps may occur */
+                       child->ptrace |= PT_DTRACE;
+               }
+               tmp = get_stack_long(child, EFL_OFFSET) | TRAP_FLAG;
+               put_stack_long(child, EFL_OFFSET, tmp);
+               set_tsk_thread_flag(child, TIF_SINGLESTEP);
                child->exit_code = data;
                /* give it a chance to run. */
                wake_up_process(child);
                ret = 0;
                break;
+       }
 
        case PTRACE_DETACH:
                /* detach a process that was attached. */
@@ -598,7 +482,7 @@ asmlinkage long sys_ptrace(long request, long pid, unsigned long addr, long data
                        ret = -EIO;
                        break;
                }
-               set_stopped_child_used_math(child);
+               child->used_math = 1;
                ret = set_fpregs(child, (struct user_i387_struct __user *)data);
                break;
        }
@@ -639,32 +523,20 @@ static void syscall_trace(struct pt_regs *regs)
 
 asmlinkage void syscall_trace_enter(struct pt_regs *regs)
 {
-       /* do the secure computing check first */
-       secure_computing(regs->orig_rax);
+       if (unlikely(current->audit_context))
+               audit_syscall_entry(current, regs->orig_rax,
+                                   regs->rdi, regs->rsi,
+                                   regs->rdx, regs->r10);
 
        if (test_thread_flag(TIF_SYSCALL_TRACE)
            && (current->ptrace & PT_PTRACED))
                syscall_trace(regs);
-
-       if (unlikely(current->audit_context)) {
-               if (test_thread_flag(TIF_IA32)) {
-                       audit_syscall_entry(current, AUDIT_ARCH_I386,
-                                           regs->orig_rax,
-                                           regs->rbx, regs->rcx,
-                                           regs->rdx, regs->rsi);
-               } else {
-                       audit_syscall_entry(current, AUDIT_ARCH_X86_64,
-                                           regs->orig_rax,
-                                           regs->rdi, regs->rsi,
-                                           regs->rdx, regs->r10);
-               }
-       }
 }
 
 asmlinkage void syscall_trace_leave(struct pt_regs *regs)
 {
        if (unlikely(current->audit_context))
-               audit_syscall_exit(current, AUDITSC_RESULT(regs->rax), regs->rax);
+               audit_syscall_exit(current, regs->rax);
 
        if ((test_thread_flag(TIF_SYSCALL_TRACE)
             || test_thread_flag(TIF_SINGLESTEP))